An in depth study of the biology of lung cancer has been conducted for some years, in an effort to develop newer, more rationale methods for the treatment of lung cancer. In the last two years, colon carcinoma has been included in this study. Major new findings are as follows: With the use of defined media, approximately 60% of small cell lung cancers (SCLC) and 40% of non-SCLC can be established as long term cultures. Using a medium devised for the growth of lung adenocarcinoma (LA4), approximately 50% of colon carcinomas can be cultured. By these techniques we have established multiple lines from all the major and many of the rare subtypes of lung and colon cancers. In particular, we have studied neuroendocrine (NE) differentiation in these tumors. NE markers studied included dopa decarboxylase, neuron specific enolase, brain isoenzyme of creatine kinase, dense core granules, chromogranin and peptide production. In lung cancers, all classic SCLC, carcinoids, and about 10% of non-SCLC tumors and lines expressed the entire range of markers. Variant SCLC lines expressed certain markers only. Of interest, about 50% of colon tumors and 70% of cell lines expressed multiple NE markers, but never the entire program. These studies indicate that NE differentiation is one of several pathways available to the stem cells in the bronchial and colonic mucosae, and that all bronchial and colonic tumors are interrelated (respectively). Colonic cancer cell lines (and bronchial adenocarcinomas to a lesser extent) express CEA and glycoproteins associated with mucus (TAG 72, CA 19-9 and Lewis blood group). We have compared various assays for in vitro drug testing (clonogenic, dye exclussion, MTT) and found them comparable. We are in the process of testing our cell lines to determine their sensitivity patterns. Studies with oncogene amplification and expression in lung cancers and cell lines are presented elsewhere. Of interest, all colonic cancer lines express c-myc mRNA, and one line is greatly amplified.